Course selection decision making is an extremely tedious task that needs to consider course prerequisites, degree requirements, class schedules, as well as the student's preferences and constraints. As a result, students often make short term decisions based on imprecise information without deep understanding of the longer-term impact on their education goal and in most cases without good understanding of the alternative options. In this paper, we introduce CourseNavigator, a new course exploration service that attempts to address the course exploration challenge. Our service identifies all possible course selection options for a given academic period, referred to as learning paths, that can meet the student's customized goals and constraints. CourseNavigator offers a suite of learning path generation algorithms designed to meet a range of course exploration end-goals such as learning paths for a given period and desired degree as well as the highest ranked paths based on user-defined ranking functions. Our techniques rely on a graph-search algorithm for enumerating candidate learning paths and employ a number of strategies (i.e., early detection of dead-end paths, limiting the exploration to strategic course selections) for improving the exploration efficiency.
{"title":"CourseNavigator: interactive learning path exploration","authors":"Zhan Li, Olga Papaemmanouil, G. Koutrika","doi":"10.1145/2948674.2948676","DOIUrl":"https://doi.org/10.1145/2948674.2948676","url":null,"abstract":"Course selection decision making is an extremely tedious task that needs to consider course prerequisites, degree requirements, class schedules, as well as the student's preferences and constraints. As a result, students often make short term decisions based on imprecise information without deep understanding of the longer-term impact on their education goal and in most cases without good understanding of the alternative options. In this paper, we introduce CourseNavigator, a new course exploration service that attempts to address the course exploration challenge. Our service identifies all possible course selection options for a given academic period, referred to as learning paths, that can meet the student's customized goals and constraints. CourseNavigator offers a suite of learning path generation algorithms designed to meet a range of course exploration end-goals such as learning paths for a given period and desired degree as well as the highest ranked paths based on user-defined ranking functions. Our techniques rely on a graph-search algorithm for enumerating candidate learning paths and employ a number of strategies (i.e., early detection of dead-end paths, limiting the exploration to strategic course selections) for improving the exploration efficiency.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123854685","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Data exploration encompasses a variety of interaction types and data functionality, such as search, data analysis, curation, constraint satisfaction, data mining, and visualization. Data exploration naturally begins when a user is given a set of data and ends when the user extracts all information and knowledge hidden in the data. Although a plethora of systems have been developed to tackle different data exploration aspects, there is no framework devoted to it as a whole. In this paper, we claim that "any" user-data interaction is essential for data exploration and sketch a prototype with both automated and user-induced functionality.
{"title":"Data exploration: a roll call of all user-data interaction functionality","authors":"Anna Gogolou, Marialena Kyriakidi, Y. Ioannidis","doi":"10.1145/2948674.2955105","DOIUrl":"https://doi.org/10.1145/2948674.2955105","url":null,"abstract":"Data exploration encompasses a variety of interaction types and data functionality, such as search, data analysis, curation, constraint satisfaction, data mining, and visualization. Data exploration naturally begins when a user is given a set of data and ends when the user extracts all information and knowledge hidden in the data. Although a plethora of systems have been developed to tackle different data exploration aspects, there is no framework devoted to it as a whole. In this paper, we claim that \"any\" user-data interaction is essential for data exploration and sketch a prototype with both automated and user-induced functionality.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125404101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recent years have seen a surge in "self-service" business intelligence tools. These tools primarily focus on supporting decision-making by non-technical "end users", through data exploration -- the querying of data and inspection of results. Exploration, however, is only part of the story. Curation is its complement. Curation is the ability to organize data into structures that are meaningful for a particular problem domain and convenient for building further explorations upon. Curation is also the ability to modify data, as well as creating new data through rules and constraints, in order to support what-if's, forecasting, and planning for the future. Exploration and curation often need to interleave in the decision-making process of an end-user. In this talk, we discuss the LogicBlox Modeler, a unifying environment that provides support for both exploration and curation. We motivate the need for a unifying environment through applications in government, major financial institutions, and large global retailers. We discuss our language -- in its visual and textual representations -- that supports not only querying, but also the creation and modification of schema and data. We discuss the challenges imposed on the database runtime by the use cases of exploration and curation at scale and aspects of the LogicBlox database designed to meet these challenges.
{"title":"Unifying data exploration and curation","authors":"S. Huang","doi":"10.1145/2948674.2948680","DOIUrl":"https://doi.org/10.1145/2948674.2948680","url":null,"abstract":"Recent years have seen a surge in \"self-service\" business intelligence tools. These tools primarily focus on supporting decision-making by non-technical \"end users\", through data exploration -- the querying of data and inspection of results. Exploration, however, is only part of the story. Curation is its complement. Curation is the ability to organize data into structures that are meaningful for a particular problem domain and convenient for building further explorations upon. Curation is also the ability to modify data, as well as creating new data through rules and constraints, in order to support what-if's, forecasting, and planning for the future. Exploration and curation often need to interleave in the decision-making process of an end-user. In this talk, we discuss the LogicBlox Modeler, a unifying environment that provides support for both exploration and curation. We motivate the need for a unifying environment through applications in government, major financial institutions, and large global retailers. We discuss our language -- in its visual and textual representations -- that supports not only querying, but also the creation and modification of schema and data. We discuss the challenges imposed on the database runtime by the use cases of exploration and curation at scale and aspects of the LogicBlox database designed to meet these challenges.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"443 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131691484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Domain novices learning about a new subject can struggle to find their way in large collections. Typical searching and browsing tools are better utilized if users know what to search for or browse to. Navigation systems with richer user interfaces could go beyond search and browse facilities by providing overviews and exploration features. We present Multiple Diagram Navigation (MDN) to assist domain novices by providing multiple overviews of the content matter. Rather than relying on specific types of visualizations, MDN superimposes any type of diagram or map over a collection of information resources, allowing content providers to reveal interesting perspectives of their content. Domain novices can navigate through the content in an exploratory way using three types of browsing (queries): diagram to content, diagram to diagram, and content to diagram. We present positive indications of MDN usability and usefulness we received from a preliminary user study. We also present our vision for using heuristics about diagram structures to help rank results returned by MDN queries.
{"title":"Multiple diagram navigation (MDN)","authors":"Hisham Benotman, L. Delcambre, D. Maier","doi":"10.1145/2948674.2948678","DOIUrl":"https://doi.org/10.1145/2948674.2948678","url":null,"abstract":"Domain novices learning about a new subject can struggle to find their way in large collections. Typical searching and browsing tools are better utilized if users know what to search for or browse to. Navigation systems with richer user interfaces could go beyond search and browse facilities by providing overviews and exploration features. We present Multiple Diagram Navigation (MDN) to assist domain novices by providing multiple overviews of the content matter. Rather than relying on specific types of visualizations, MDN superimposes any type of diagram or map over a collection of information resources, allowing content providers to reveal interesting perspectives of their content. Domain novices can navigate through the content in an exploratory way using three types of browsing (queries): diagram to content, diagram to diagram, and content to diagram. We present positive indications of MDN usability and usefulness we received from a preliminary user study. We also present our vision for using heuristics about diagram structures to help rank results returned by MDN queries.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121064671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Miao Cheng, Anand Sriramulu, S. Muralidhar, B. T. Loo, Laura Huang, Po-Ling Loh
Crowdfunding is a recent financing phenomenon that is gaining wide popularity as a means for startups to raise seed funding for their companies. This paper presents our initial results at understanding this phenomenon using an exploratory data driven approach. We have developed a big data platform for collecting and managing data from multiple sources, including company profiles (CrunchBase and AngelList) and social networks (Facebook and Twitter). We describe our data collection process that allows us to gather data from diverse sources at high throughput. Using Spark as our analysis tool, we study the impact of social engagement on startup fund raising success. We further define novel metrics that allow us to quantify the behavior of investors to follow and make investment decisions as communities rather than individuals. Finally, we explore visualization techniques that allow us to visualize communities of investors that make decisions in a close-knit fashion vs looser communities where investors largely make independent decisions. We conclude with a discussion on our ongoing research on causality analysis and new community detection algorithms.
{"title":"Collection, exploration and analysis of crowdfunding social networks","authors":"Miao Cheng, Anand Sriramulu, S. Muralidhar, B. T. Loo, Laura Huang, Po-Ling Loh","doi":"10.1145/2948674.2948679","DOIUrl":"https://doi.org/10.1145/2948674.2948679","url":null,"abstract":"Crowdfunding is a recent financing phenomenon that is gaining wide popularity as a means for startups to raise seed funding for their companies. This paper presents our initial results at understanding this phenomenon using an exploratory data driven approach. We have developed a big data platform for collecting and managing data from multiple sources, including company profiles (CrunchBase and AngelList) and social networks (Facebook and Twitter). We describe our data collection process that allows us to gather data from diverse sources at high throughput. Using Spark as our analysis tool, we study the impact of social engagement on startup fund raising success. We further define novel metrics that allow us to quantify the behavior of investors to follow and make investment decisions as communities rather than individuals. Finally, we explore visualization techniques that allow us to visualize communities of investors that make decisions in a close-knit fashion vs looser communities where investors largely make independent decisions. We conclude with a discussion on our ongoing research on causality analysis and new community detection algorithms.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133812457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Fernandez, Ziawasch Abedjan, S. Madden, M. Stonebraker
With thousands of data sources spread across multiple databases and data lakes, modern organizations face a data discovery challenge. Analysts spend more time finding relevant data to answer the questions at hand than analyzing it. In this paper we introduce a data discovery system that facilitates locating relevant data among thousands of data sources. We represent data sources succinctly through signatures, and then create search paths that permit quick execution of a set of data discovery primitives used for finding relevant data. We have built a prototype that is being used to solve data discovery challenges of two big organizations.
{"title":"Towards large-scale data discovery: position paper","authors":"R. Fernandez, Ziawasch Abedjan, S. Madden, M. Stonebraker","doi":"10.1145/2948674.2948675","DOIUrl":"https://doi.org/10.1145/2948674.2948675","url":null,"abstract":"With thousands of data sources spread across multiple databases and data lakes, modern organizations face a data discovery challenge. Analysts spend more time finding relevant data to answer the questions at hand than analyzing it. In this paper we introduce a data discovery system that facilitates locating relevant data among thousands of data sources. We represent data sources succinctly through signatures, and then create search paths that permit quick execution of a set of data discovery primitives used for finding relevant data. We have built a prototype that is being used to solve data discovery challenges of two big organizations.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114072155","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mirjana Pavlovic, Eleni Tzirita Zacharatou, Darius Sidlauskas, T. Heinis, A. Ailamaki
Advances in data acquisition---through more powerful supercomputers for simulation or sensors with better resolution---help scientists tremendously to understand natural phenomena. At the same time, however, it leaves them with a plethora of data and the challenge of analysing it. Ingesting all the data in a database or indexing it for an efficient analysis is unlikely to pay off because scientists rarely need to analyse all data. Not knowing a priori what parts of the datasets need to be analysed makes the problem challenging. Tools and methods to analyse only subsets of this data are rather rare. In this paper we therefore present Space Odyssey, a novel approach enabling scientists to efficiently explore multiple spatial datasets of massive size. Without any prior information, Space Odyssey incrementally indexes the datasets and optimizes the access to datasets frequently queried together. As our experiments show, through incrementally indexing and changing the data layout on disk, Space Odyssey accelerates exploratory analysis of spatial data by substantially reducing query-to-insight time compared to the state of the art.
{"title":"Space odyssey: efficient exploration of scientific data","authors":"Mirjana Pavlovic, Eleni Tzirita Zacharatou, Darius Sidlauskas, T. Heinis, A. Ailamaki","doi":"10.1145/2948674.2948677","DOIUrl":"https://doi.org/10.1145/2948674.2948677","url":null,"abstract":"Advances in data acquisition---through more powerful supercomputers for simulation or sensors with better resolution---help scientists tremendously to understand natural phenomena. At the same time, however, it leaves them with a plethora of data and the challenge of analysing it. Ingesting all the data in a database or indexing it for an efficient analysis is unlikely to pay off because scientists rarely need to analyse all data. Not knowing a priori what parts of the datasets need to be analysed makes the problem challenging. Tools and methods to analyse only subsets of this data are rather rare. In this paper we therefore present Space Odyssey, a novel approach enabling scientists to efficiently explore multiple spatial datasets of massive size. Without any prior information, Space Odyssey incrementally indexes the datasets and optimizes the access to datasets frequently queried together. As our experiments show, through incrementally indexing and changing the data layout on disk, Space Odyssey accelerates exploratory analysis of spatial data by substantially reducing query-to-insight time compared to the state of the art.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131509976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With so many advances in machine learning recently, it's not unreasonable to ask: why aren't my recommendations perfect by now? Aish provides a walkthrough of the open problems in the area of recommender systems, especially as they apply to Netflix's personalization and recommender algorithms. He also provides a brief overview of recommender systems, and sketches out some tentative solutions for the problems he presents.
{"title":"Why would you recommend me that!?","authors":"Aish Fenton","doi":"10.1145/2948674.2948681","DOIUrl":"https://doi.org/10.1145/2948674.2948681","url":null,"abstract":"With so many advances in machine learning recently, it's not unreasonable to ask: why aren't my recommendations perfect by now? Aish provides a walkthrough of the open problems in the area of recommender systems, especially as they apply to Netflix's personalization and recommender algorithms. He also provides a brief overview of recommender systems, and sketches out some tentative solutions for the problems he presents.","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131882385","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","authors":"","doi":"10.1145/2948674","DOIUrl":"https://doi.org/10.1145/2948674","url":null,"abstract":"","PeriodicalId":165112,"journal":{"name":"Proceedings of the Third International Workshop on Exploratory Search in Databases and the Web","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125034828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：